Understanding the agglomeration behavior of nickel laterite and gold ores using statistical design of experiments

The drum agglomeration of nickel laterite and gold ores has been optimized through the design of experiments (DOE) using a Taguchi L16 (45) orthogonal array to determine the optimum conditions for maximizing average agglomerate size and minimizing the amount of fines. The effects of controllable operating factors including moisture content (nickel laterite ore: 34-37%; gold ore: 7-10%), retention time (2-3.5 min), drum speed (15-45% critical speed), drum load (nickel laterite ore: 8-32 %; gold ore: 6-22%) and acid concentration (150-600 g/L) on the performance of the agglomeration process were studied. For nickel laterite ore, maximum average agglomerate size and minimum percent fines (-1 mm) occurred under the following conditions: drum load (23.7%), moisture (36.5%), time (3 min), drum speed (30% critical speed) and acid concentration (150 g/L). Under the studied nickel laterite ore conditions, the most effective parameters for maximizing average agglomerate size and minimizing the amount of fines were found to be drum load and acid concentration, respectively. Drum speed had a statistically significant effect on minimizing the amount of fines. Maximum average agglomerate size and minimum percent fines (-1 mm) for gold ore occurred under the following conditions: drum load (19.3%), moisture (8.5%), time (2 min 15 s) and drum speed (40% critical). The most significant factors for maximizing average agglomerate size and minimizing the amount of fines for gold ore were found to be drum load, time and moisture. Paper number MMP-12-100. Original manuscript submitted December 2012. Revised manuscript accepted for publication March 2013. Discussion of this peer-reviewed and approved paper is invited and must be submitted to SME Publications Dept. prior to August 31, 2014. Copyright 2014, Society for Mining, Metallurgy & Exploration Inc. Introduction Several heap leach operations have experienced problems in terms of poor recovery due to percolation issues caused by low-grade complex ores, tailings and clayey deposits. Poor percolation can lead to low metal extraction due to solution channeling or the development of impermeable/dead zones within the heap (Dhawan et al., 2012; Kappes, 2005). Improper heap building practices such as the use of loose agglomerates and inadequate attention toward the presence of clay minerals have been among the main reasons for percolation issues. During the transport of ore material, severe segregation of the material can occur. To overcome percolation problems, a major improvement was made through the introduction of agglomeration prior to ore placement. If the ore particles and agglomerates are of similar size, segregation can be avoided to a great extent (Dhawan et al., 2013; Kinard and Schweizer, 1987; McClelland and van Zyl, 1988). Agglomeration improves the uniform percolation of solution through ore heaps and is applicable to many ores, wastes and milled tailings (Bouffard, 2005; Dhawan et al., 2013). Manning and Kappes (2005) reported agglomeration/ stacking accounts for ~14% of the total heap leaching operating cost. The cost of binder is a primary contributor toward the total cost. Also, the lack of consistent quality control tests for agglomerate often leads to operational problems. The details are mentioned elsewhere (Dhawan et al., 2013). Considering the significant contribution of the agglomeration/stacking step toward the total heap leach operating cost, it is appropriate to study the agglomeration behavior and its optimization for two different ores. The reason to study the same parameters for different ores is the fact that significant variability in the agglomeration behavior of different ores has been reported to exist in industrial operations. It is well understood that a wide feed particle size distribution (PSD) is not ideal for consistent high-quality agglomerates, when it is known that agglomerates must undergo mechanical